GRAML: Goal Recognition As Metric Learning

Matan Shamir; Reuth Mirsky

REVIEW 3 major objections 5 minor 35 references

GRAML: Goal Recognition As Metric Learning

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that goal recognition can be cast as metric learning, so a newly introduced goal is recognized from a single example trace in a learned embedding space.

desk verdict A real, if narrowly bounded, contribution to online dynamic goal recognition: the metric-learning idea works and is evaluated on held-out goals, but the abstract overstates runtime and the base-goal coverage boundary needs to be pinned down. read the letter →

arxiv 2505.03941 v2 pith:GD3EFJ4J submitted 2025-05-06 cs.AI

Matan Shamir , Reuth Mirsky This is my paper · ORCID

classification cs.AI

keywords goalrecognitionmetriclearningonlinedynamicone-shotadaptationSiamesenetworkLSTMembeddingsgoal-conditionedreinforcement

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper aims to turn goal recognition—inferring what an agent is trying to achieve from the actions it has taken—into a metric-learning problem. Its claim is that a recurrent network trained on traces that lead to a fixed set of base goals learns a distance in embedding space in which traces to the same goal are close and traces to different goals are far apart. Once that metric exists, a newly introduced goal is handled by embedding a single example trace for that goal and comparing new observations against it, so no retraining per goal is needed. The authors evaluate this on discrete and continuous environments with partial, noisy, and suboptimal observations, and report one-shot adaptation speeds that avoid the policy-retraining cost of previous approaches. If the claim holds, dynamic environments whose goal sets change over time gain a recognition method whose adaptation phase is a cheap library update rather than a full learning problem.

What carries the argument

The load-bearing object is the embedding space $\mathbb{R}^k$ produced by the Siamese LSTM. The similarity function is $e^{-\sum_{i} |v_1[i]-v_2[i]|}$ (Eq. 1), trained by binary cross-entropy on same-goal and different-goal trace pairs (Eq. 2). Inference is nearest-neighbor in that space with averaging over $s$ example embeddings per goal (Eq. 3). The mechanism's work is to convert a goal recognition problem into a nearest-neighbor lookup, so that adapting to new goals only requires embedding their example traces through the fixed LSTM.

What would settle it

Train GRAML on base goals restricted to one corner of a gridworld, then introduce a goal in the opposite corner and measure one-shot accuracy on partial traces; if accuracy stays at chance while base-goal traces are still recognized, the metric-transfer claim is falsified.

Watch

Extended reading notes

Core claim

GRAML's central discovery is that the recognition step can be decoupled from goal identity: instead of learning what each goal looks like, learn a metric over the space of observation traces. Concretely, a Siamese network with two shared-weight LSTM branches is trained with binary cross-entropy to produce embeddings $v$ such that the exponential of negative Manhattan distance (Eq. 1) is near 1 for same-goal pairs and near 0 for different-goal pairs. At inference, each active goal is represented by a small library of embeddings of example traces, and a query trace is assigned to the goal whose average similarity is maximal (Eq. 3). The paper claims this metric transfers to goals unseen in training, which is what makes one-shot adaptation possible: one example trace per new goal is enough to locate that goal in the learned space.

Load-bearing premise

The learned metric transfers from the base goals used in training to any new goal introduced later, so if the base goals do not cover the goal space, an unseen goal's traces can land anywhere in embedding space and recognition fails.

Editorial extensions

If this is right

Adding a goal to the active set costs one example trace per new goal: no retraining of the recognition network and no policy learning for that goal.
Continuous domains need no discretization, because the LSTM maps continuous state-action pairs into a fixed embedding space rather than predicting in state space.
Recognition accuracy is expected to rise with the number of base goals and fall as the number of simultaneous active goals grows, since more goals create more confusable traces.
The same learned metric can serve repeated goal-set changes in an online setting: each new set of goals only triggers the adaptation phase, not a new domain learning phase.
Larger example libraries per goal improve accuracy at linearly increasing adaptation and runtime cost, giving a tunable accuracy-speed trade-off.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Editorial inference: if the transfer claim is right, the same embedding metric could be reused across many different goal sets in the same domain, making GRAML a pretrained 'goal-space prior' that is updated by appending embeddings rather than by re-optimizing.
Editorial inference: the method implicitly assumes that distinct goals are geometrically separable in trajectory space; goals that share all near-optimal paths with another goal would remain indistinguishable unless the library contains disambiguating traces, a limitation the paper does not analyze.
Editorial inference: the embedding space itself could be used to measure similarity between goals, which would give an automatic way to choose base goals that cover the goal space—directly addressing the base-goal selection issue the paper leaves open.
Editorial inference: a straightforward test of the mechanism is to probe whether the learned metric is approximately monotone with true plan distance; if it is, the same model could also serve as a planner heuristic, not just a recognizer.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The GRAML paper is a genuine step forward for goal recognition under changing goal sets: it reformulates the problem as metric learning, shows one-shot adaptation on held-out goals, and tests on six environments covering discrete and continuous control. The core idea is simple and transferable: train a Siamese LSTM so traces to the same goal embed near each other, then at inference just embed one reference trace per new goal and do nearest-neighbor. That is actually new for this problem, and the held-out evaluation is the right kind of evidence, so the central claim is not circular.

The paper also does a few things well. It introduces two variants, BG and GC, the goal-conditioned one is a natural fit, and the benchmark suite (Minigrid, PointMaze, Parking, Panda) gives the field a common testbed that was missing for ODGR. The ablation on number of base goals is useful, and the explicit discussion of observability and suboptimal traces is honest. Section 6 is unusually candid about the base-goal coverage limitation.

The soft spots are real but not fatal. The abstract's 'runtime improvements' line is too broad: BG-GRAML's zero adaptation time comes from expert-provided sequences, and the paper later admits that planning for the library takes about four times less than retraining. That is still an improvement, but the framing in the abstract overstates it. A bigger issue is the missing comparison to GRNet, the closest dynamic-goal baseline. The paper mentions GRNet's fluent enumeration weakness, but never runs it, so we don't know if the added complexity buys accuracy. The stress-test concern about base-goal coverage is legitimate: the loss only constrains distances among base goals, and Section 6 concedes that uncovered state-space regions may not be recognized. The held-out evaluation should report how far the active goals lie from the base-goal distribution. That is the main missing number.

Overall, the paper shows clear thinking and honest engagement with its own limitations. I would send it to peer review, but the authors should be asked to (1) report base-goal distances/coverage, (2) add a GRNet comparison where feasible, and (3) tone down the abstract's runtime claim. For a reader working on plan or goal recognition, this is worth reading now; for others, it is a solid application of metric learning with a clean ODGR formulation.

Referee Report

3 major / 5 minor

Summary. The paper introduces GRAML, a goal-recognition method that casts recognition as metric learning. A Siamese LSTM is trained so that observation traces leading to the same goal are embedded closer to each other than traces leading to different goals; at inference, a new observation is classified by nearest-neighbor comparison against one or more reference traces per candidate goal. Two variants are proposed: BG-GRAML, which trains separate agents for a set of base goals, and GC-GRAML, which trains a single goal-conditioned policy. The method is evaluated on six discrete and continuous environments, under partial consecutive and non-consecutive observations, stochastic suboptimal traces, and held-out goals not used during training. The paper reports accuracy comparable to or better than the GR-as-RL baselines GRAQL and DRACO, with reduced goal-adaptation time, and analyzes the effect of the number of base and active goals.

Significance. If the central claim holds—that a metric learned on a set of base goals transfers to newly introduced goals from a single reference trace—GRAML would be a useful contribution to data-driven goal recognition, particularly for online dynamic goal recognition in continuous domains. The paper has notable strengths: it evaluates on six environments spanning discrete and continuous state/action spaces; it tests partial, non-consecutive, and stochastic suboptimal observations; and it uses held-out goals in the evaluation, which is the right kind of evidence against circularity. The proposed ODGR phase decomposition is also clearly articulated. However, the generalization claim is not yet fully delimited: the relationship between the base-goal set and the held-out active goals is not measured, and the paper itself concedes that base-goal selection may be crucial. The evaluation also omits GRNet, the closest model-free dynamic-goal baseline, and the runtime comparison assigns zero cost to expert-provided reference sequences. These issues are addressable but currently leave the scope of the central claim and the speed advantage not fully established.

major comments (3)

[Section 3.1.1, Section 4.3, Section 6] The central claim that the learned metric transfers to newly introduced goals is not yet delimited. The training loss in Eqs. (1)-(2) only constrains distances among traces to the chosen base goals, and Eq. (3) assumes that reference traces for previously unseen goals are embedded near same-goal traces and far from other-goal traces. The held-out evaluation in Section 4.3 never reports the relationship between the base-goal set and the active goal set G0, such as distance in state space, overlap with the base-goal distribution, or difficulty of the goals. Section 6 explicitly concedes that base-goal choice 'may be crucial for the model's success' and that uncovered areas of the state space may not be recognized correctly. The paper therefore demonstrates one-shot transfer for its specific base-goal choices, but it does not establish the boundary of the claimed generalization. Please report the coverage of the base goals relative to G0 (e.g., distance distributions, or probes into regions deliberately far from all base goals), or restrict the claim to goal sets sampled near the base-goal distribution.
[Section 5, Section 4.3] The closest dynamic-goal baseline, GRNet, is discussed in Related Work but is not included in the accuracy or runtime comparisons. GRNet is described as generalizing to any set of goals without additional learning, which is exactly the capability GRAML claims to improve. Without a comparison to GRNet, the phrases 'state-of-the-art GR' and the speed/runtime improvements are not demonstrated against the most relevant method. Please add GRNet to at least the discrete benchmarks, or provide a concrete incompatibility justification with evidence rather than a qualitative remark about fluent enumeration.
[Section 4.3 (runtime paragraph)] The runtime comparison credits BG-GRAML with G = 0 when expert sequences are provided, while the same paragraph reports 75 seconds per goal for MCTS planning and 18 minutes for a five-goal problem. The phase accounting omits the cost of producing the reference library and the domain-learning cost of training base-goal agents, which can dominate in BG-GRAML. Please report a complete wall-clock breakdown for all phases—domain learning, library generation, goal adaptation, and inference—on the same hardware, and state whether expert/oracle time is included or excluded. Without this, the speed advantage is partly an artifact of zero-cost assumptions.

minor comments (5)

[Section 4.2, Figure 5] The comparison between GC-GRAML and BG-GRAML is confounded because the two variants use different numbers of base goals (|G| = 20 for GC-GRAML and |G| = 5 for BG-GRAML). The conclusion that GC-GRAML benefits from the variety of sequences is therefore not isolated; please report both variants with matched |G| or explicitly analyze the effect.
[Table 1] The accuracy table reports means and standard deviations but no significance tests or confidence intervals. Several claimed improvements appear to be within one standard deviation; please add statistical testing or clearly state which differences are reliable.
[Section 4.3 and Figure 1] The notation 'G = 0' in the runtime paragraph is confusing because G is also used for the goal set and for the goals-adaptation interval; please use a distinct symbol such as T_G or t_adapt.
[Section 3.1.1] There is a typo: 'a single GCRL agent is trained on ¯G, which can also be defined as a continuum or discrete set. The sample generation is also done similarly, but in this case, a single goal-conditioned policy generates the traces.' The phrase 'buts instead of multiple agents' appears garbled and should be corrected.
[General] The repository link is promised only post-acceptance, so the current manuscript does not provide the artifacts needed to reproduce the experiments. Please include an anonymized repository link or a detailed appendix with hyperparameters and environment configurations.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation; held-out goal transfer is real. Minor self-citation of the ODGR problem definition is not load-bearing.

full rationale

GRAML's central claim—that a metric learned over base-goal traces supports one-shot recognition of newly introduced goals—is not circular. The training objective (Eqs. 1–2) is supervised only by same/different labels over the chosen base goals \bar{G}; the active goals G0 are held out and enter only at inference through the reference traces in Eq. (3) and the nearest-goal argmax. Section 4.3 and Figure 4 explicitly evaluate on goals that 'differ from the set of base goals from the domain learning phase,' so the reported accuracy is a genuine held-out-transfer measurement rather than a refit. The evaluation is also anchored to external baselines (GRAQL, DRACO) and to Gymnasium-based environments, so the empirical content is not self-confirming. The only identifiable self-citation is Definition 3 (ODGR), attributed to Shamir et al. 2024; that definition frames the problem but does not determine the embedding or supply any fitted value, so it is not load-bearing. Section 6 concedes that base-goal selection 'may be crucial for the model's success' and that uncovered state-space regions may not be recognized; this is an honest scope limitation on how far the generalization claim extends, not evidence that the demonstrated transfer reduces to the inputs. No equation or fitted parameter is renamed as a prediction. Accordingly, there are no circular steps; the low score reflects only the minor, non-load-bearing self-citation of the problem formulation.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central claim rests on one transfer assumption (base goals to new goals), one capacity hyperparameter (number of base goals), and the library size. No new physical entities are introduced; the ODGR formulation is carried over from the authors' prior work.

free parameters (2)

Number of base goals |G_bar| = 5 for BG-GRAML, 20 for GC-GRAML
Chosen by hand; Figure 6 shows accuracy increases with base goals, so this is an ad hoc capacity choice.
Library size s = 1 in the main accuracy comparison
Hyperparameter controlling how many sequences per goal are embedded; paper notes accuracy increases with s at the cost of runtime.

assumptions (3)

domain assumption Traces to the same goal are closer in the learned embedding than traces to different goals
This is the metric-learning objective itself, enforced during training via Eq. (1)-(2); the paper assumes it holds for new goals.
domain assumption The embedding trained on base goals generalizes to any new goal in the same domain
This is the key transfer premise; if it fails, one-shot adaptation fails. The paper's Section 6 admits base goal selection is critical.
domain assumption Goal recognition can be cast as nearest-neighbor retrieval among goal embeddings
The inference rule Eq. (3) assumes the closest embedding corresponds to the true goal, which holds only if the learned metric is well calibrated.

how reviews work

0 comments

Cite this review

Pith. "Pith review of GRAML: Goal Recognition As Metric Learning." pith.science (2026). https://pith.science/paper/GD3EFJ4J

@misc{pith2026250503941,
  author       = {Pith},
  title        = {Pith review of: GRAML: Goal Recognition As Metric Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GD3EFJ4J}},
  note         = {Machine review of arXiv:2505.03941}
}

read the original abstract

Goal Recognition (GR) is the problem of recognizing an agent's objectives based on observed actions. Recent data-driven approaches for GR alleviate the need for costly, manually crafted domain models. However, these approaches can only reason about a pre-defined set of goals, and time-consuming training is needed for new emerging goals. To keep this model-learning automated while enabling quick adaptation to new goals, this paper introduces GRAML: Goal Recognition As Metric Learning. GRAML uses a Siamese network to treat GR as a deep metric learning task, employing an RNN that learns a metric over an embedding space, where the embeddings for observation traces leading to different goals are distant, and embeddings of traces leading to the same goals are close. This metric is especially useful when adapting to new goals, even if given just one example observation trace per goal. Evaluated on a versatile set of environments, GRAML shows speed, flexibility, and runtime improvements over the state-of-the-art GR while maintaining accurate recognition.

Figures

Figures reproduced from arXiv: 2505.03941 by the authors.

**Figure 2.** GRAML’s implementation for the different phases of an ODGR problem. The numbers at the top left of each box reflect the section [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 4.** Confusion matrices for plan similarity and recognition [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

Figures from the paper (1 more)

**Figure 6.** Figure 6: BG-GRAML’s accuracy as a factor of the number of newly [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 30 canonical work pages

[1]

Goal recognition as reinforcement learning

[Amado et al., 2022] Leonardo Amado, Reuth Mirsky, and Felipe Meneguzzi. Goal recognition as reinforcement learning. In The AAAI Conference on Artificial Intelli- gence, volume 36, pages 9644–9651,

work page 2022
[4]

Fast and complete symbolic plan recognition

[Avrahami-Zilberbrand and Kaminka, 2005] Dorit Avrahami-Zilberbrand and Gal A Kaminka. Fast and complete symbolic plan recognition. In IJCAI, pages 653–658,

work page 2005
[8]

Goal recognition as a deep learning task: The grnet approach

[Chiari et al., 2023] Mattia Chiari, Alfonso Emilio Gerevini, Francesco Percassi, Luca Putelli, Ivan Serina, and Matteo Olivato. Goal recognition as a deep learning task: The grnet approach. In The International Conference on Au- tomated Planning and Scheduling, volume 33, pages 560– 568,

work page 2023
[9]

Deep localized metric learning

[Duan et al., 2018] Yueqi Duan, Jiwen Lu, Jianjiang Feng, and Jie Zhou. Deep localized metric learning. IEEE Transactions on Circuits and Systems for Video Technol- ogy, 28(10):2644–2656,

work page 2018
[11]

General dynamic goal recognition

[Elhadad and Mirsky, 2025] Osher Elhadad and Reuth Mirsky. General dynamic goal recognition. arXiv preprint arXiv:2505.09737,

work page arXiv 2025
[13]

Integration of planning with recog- nition for responsive interaction using classical planners

[Freedman and Zilberstein, 2017] Richard Freedman and Shlomo Zilberstein. Integration of planning with recog- nition for responsive interaction using classical planners. In The AAAI Conference on Artificial Intelligence , volume 31,

work page 2017
[15]

panda- gym: Open-Source Goal-Conditioned Environments for Robotic Learning

[Gallou´edec et al., 2021] Quentin Gallou ´edec, Nicolas Cazin, Emmanuel Dellandr ´ea, and Liming Chen. panda- gym: Open-Source Goal-Conditioned Environments for Robotic Learning. 4th Robot Learning Workshop: Self-Supervised and Lifelong Learning at NeurIPS,

work page 2021
[17]

A probabilistic plan recognition algo- rithm based on plan tree grammars

[Geib and Goldman, 2009] Christopher W Geib and Robert P Goldman. A probabilistic plan recognition algo- rithm based on plan tree grammars. Artificial Intelligence, 173(11):1101–1132,

work page 2009

Show all 35 references

[18]

Learning combinatory categorial grammars for plan recognition

[Geib and Kantharaju, 2018] Christopher Geib and Pavan Kantharaju. Learning combinatory categorial grammars for plan recognition. In The AAAI Conference on Artificial Intelligence, volume 32,

2018
[19]

An environment for au- tonomous driving decision-making

[Leurent, 2018] Edouard Leurent. An environment for au- tonomous driving decision-making. https://github.com/ eleurent/highway-env,

2018
[20]

Goal-conditioned reinforcement learning: Prob- lems and solutions

[Liu et al., 2022] Minghuan Liu, Menghui Zhu, and Weinan Zhang. Goal-conditioned reinforcement learning: Prob- lems and solutions. arXiv preprint arXiv:2201.08299 ,

2022 arXiv
[21]

Parc: A plan and activity recognition com- ponent for assistive robots

[Massardi et al., 2020] Jean Massardi, Mathieu Gravel, and ´Eric Beaudry. Parc: A plan and activity recognition com- ponent for assistive robots. In 2020 IEEE International Conference on Robotics and Automation (ICRA) , pages 3025–3031. IEEE,

2020
[22]

A survey on goal recognition as planning

[Meneguzzi and Fraga Pereira, 2021] Felipe Meneguzzi and Ramon Fraga Pereira. A survey on goal recognition as planning. In International Joint Conference on Artificial Intelligence, IJCAI-21, pages 4524–4532, 8

2021
[23]

Deep learning-based goal recognition in open-ended digital games

[Min et al., 2014] Wookhee Min, Eun Ha, Jonathan Rowe, Bradford Mott, and James Lester. Deep learning-based goal recognition in open-ended digital games. InThe AAAI Conference on Artificial Intelligence and Interactive Digi- tal Entertainment, volume 10, pages 37–43,

2014
[25]

Siamese recurrent architectures for learning sentence similarity

[Mueller and Thyagarajan, 2016] Jonas Mueller and Aditya Thyagarajan. Siamese recurrent architectures for learning sentence similarity. The AAAI Conference on Artificial In- telligence, 30(1), Mar

2016
[28]

[Rakthanmanon et al., 2012] Thanawin Rakthanmanon, Bil- son J. L. Campana, Abdullah Al Mueen, Gustavo E. A. P. A. Batista, M. Brandon Westover, Qiang Zhu, Jesin Zakaria, and Eamonn J. Keogh. Searching and mining trillions of time series subsequences under dynamic time warping....

2012
[30]

Goal recognition over pomdps: Inferring the in- tention of a pomdp agent

[Ramırez and Geffner, 2011] Miquel Ramırez and Hector Geffner. Goal recognition over pomdps: Inferring the in- tention of a pomdp agent. In IJCAI, pages 2009–2014. IJCAI/AAAI,

2011
[32]

Proactive robotic assistance via theory of mind

[Shvo et al., 2022] Maayan Shvo, Ruthrash Hari, Ziggy O’Reilly, Sophia Abolore, Sze-Yuh Nina Wang, and Sheila A McIlraith. Proactive robotic assistance via theory of mind. In International Conference on Intelligent Robots and Systems (IROS), pages 9148–9155. IEEE,

2022
[33]

Plan recognition as planning revisited

[Sohrabi et al., 2016] Shirin Sohrabi, Anton V Riabov, and Octavian Udrea. Plan recognition as planning revisited. In IJCAI, pages 3258–3264. New York, NY ,

2016
[34]

Fast and ac- curate data-driven goal recognition using process mining techniques

[Su et al., 2023] Zihang Su, Artem Polyvyanyy, Nir Lipovet- zky, Sebastian Sardi˜na, and Nick van Beest. Fast and ac- curate data-driven goal recognition using process mining techniques. Artificial Intelligence, 323:103973,

2023
[35]

Hardness-aware deep metric learn- ing

[Zheng et al., 2019] Wenzhao Zheng, Zhaodong Chen, Ji- wen Lu, and Jie Zhou. Hardness-aware deep metric learn- ing. In Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition, pages 72–81,

2019
[2005]

Baker, Rebecca Saxe, and Joshua B

[Baker et al., 2009] Chris L. Baker, Rebecca Saxe, and Joshua B. Tenenbaum. Action understanding as inverse planning. Cognition, 113(3):329–349,

2009
[2009]

Goal recognition via model- based and model-free techniques

[Borrajo et al., 2020] Daniel Borrajo, Sriram Gopalakrish- nan, and Vamsi K Potluru. Goal recognition via model- based and model-free techniques. FinPlan 2020, page 10,

2020
[2011]

Odgr: Online dynamic goal recognition

[Shamir et al., 2024] Matan Shamir, Osher Elhadad, and Reuth Mirsky. Odgr: Online dynamic goal recognition. In The Workshops of The RLC Conference,

2024
[2012]

Ramirez and H

[Ramirez and Geffner, 2009] M. Ramirez and H. Geffner. Plan recognition as planning. In International Joint Con- ference on Artificial Intelligence (IJCAI-09),

2009
[2014]

Introduction to Symbolic Plan and Goal Recognition, pages 1–120

[Mirsky et al., 2021] Reuth Mirsky, Sarah Keren, and Christopher Geib. Introduction to Symbolic Plan and Goal Recognition, pages 1–120. Number 1 in Synthesis Lectures on Artificial Intelligence and Machine Learning. Morgan and Claypool Publishers, 1 edition, January

2021
[2016]

Goal recognition using actor-critic opti- mization

[Nageris et al., 2024] Ben Nageris, Felipe Meneguzzi, and Reuth Mirsky. Goal recognition using actor-critic opti- mization. arXiv preprint arXiv:2501.01463,

2024 arXiv
[2017]

D4rl: Datasets for deep data-driven reinforcement learning

[Fu et al., 2020] Justin Fu, Aviral Kumar, Ofir Nachum, George Tucker, and Sergey Levine. D4rl: Datasets for deep data-driven reinforcement learning. arXiv preprint arXiv:2004.07219,

2020 arXiv
[2018]

Deep adversarial metric learning

[Duan et al., 2020] Yueqi Duan, Jiwen Lu, Wenzhao Zheng, and Jie Zhou. Deep adversarial metric learning. IEEE Transactions on Image Processing, 29:2037–2051,

2020
[2020]

Minigrid & miniworld: Modular & customizable reinforcement learning environments for goal-oriented tasks

[Chevalier-Boisvert et al., 2023] Maxime Chevalier- Boisvert, Bolun Dai, Mark Towers, Rodrigo de Lazcano, Lucas Willems, Salem Lahlou, Suman Pal, Pablo Samuel Castro, and Jordan Terry. Minigrid & miniworld: Modular & customizable reinforcement learning environments for goal-or...

2023 arXiv
[2021]

Model-free, model-based, and general intelligence

[Geffner, 2018] Hector Geffner. Model-free, model-based, and general intelligence. InInternational Joint Conference on Artificial Intelligence,

2018
[2022]

Robust neuro- symbolic goal and plan recognition

[Amado et al., 2023] Leonardo Amado, Ramon Fraga Pereira, and Felipe Meneguzzi. Robust neuro- symbolic goal and plan recognition. The AAAI Conference on Artificial Intelligence, 37(10):11937–11944, Jun

2023
[2023]

Classical planning in deep latent space

[Asai et al., 2022] Masataro Asai, Hiroshi Kajino, Alex Fukunaga, and Christian Muise. Classical planning in deep latent space. Journal of Artificial Intelligence Research , 74:1599–1686,

2022
[2024]

Stable-baselines3: Reliable reinforcement learning implementations

[Raffin et al., 2021] Antonin Raffin, Ashley Hill, Adam Gleave, Anssi Kanervisto, Maximilian Ernestus, and Noah Dormann. Stable-baselines3: Reliable reinforcement learning implementations. Journal of Machine Learning Research, 22(268):1–8,

2021
[2025]

Real-time online goal recogni- tion in continuous domains via deep reinforcement learn- ing

[Fang et al., 2023] Zihao Fang, Dejun Chen, Yunxiu Zeng, Tao Wang, and Kai Xu. Real-time online goal recogni- tion in continuous domains via deep reinforcement learn- ing. Entropy, 25(10):1415,

2023

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Goal recognition as reinforcement learning

[Amado et al., 2022] Leonardo Amado, Reuth Mirsky, and Felipe Meneguzzi. Goal recognition as reinforcement learning. In The AAAI Conference on Artificial Intelli- gence, volume 36, pages 9644–9651,

work page 2022

[2] [4]

Fast and complete symbolic plan recognition

[Avrahami-Zilberbrand and Kaminka, 2005] Dorit Avrahami-Zilberbrand and Gal A Kaminka. Fast and complete symbolic plan recognition. In IJCAI, pages 653–658,

work page 2005

[3] [8]

Goal recognition as a deep learning task: The grnet approach

[Chiari et al., 2023] Mattia Chiari, Alfonso Emilio Gerevini, Francesco Percassi, Luca Putelli, Ivan Serina, and Matteo Olivato. Goal recognition as a deep learning task: The grnet approach. In The International Conference on Au- tomated Planning and Scheduling, volume 33, pages 560– 568,

work page 2023

[4] [9]

Deep localized metric learning

[Duan et al., 2018] Yueqi Duan, Jiwen Lu, Jianjiang Feng, and Jie Zhou. Deep localized metric learning. IEEE Transactions on Circuits and Systems for Video Technol- ogy, 28(10):2644–2656,

work page 2018

[5] [11]

General dynamic goal recognition

[Elhadad and Mirsky, 2025] Osher Elhadad and Reuth Mirsky. General dynamic goal recognition. arXiv preprint arXiv:2505.09737,

work page arXiv 2025

[6] [13]

Integration of planning with recog- nition for responsive interaction using classical planners

[Freedman and Zilberstein, 2017] Richard Freedman and Shlomo Zilberstein. Integration of planning with recog- nition for responsive interaction using classical planners. In The AAAI Conference on Artificial Intelligence , volume 31,

work page 2017

[7] [15]

panda- gym: Open-Source Goal-Conditioned Environments for Robotic Learning

[Gallou´edec et al., 2021] Quentin Gallou ´edec, Nicolas Cazin, Emmanuel Dellandr ´ea, and Liming Chen. panda- gym: Open-Source Goal-Conditioned Environments for Robotic Learning. 4th Robot Learning Workshop: Self-Supervised and Lifelong Learning at NeurIPS,

work page 2021

[8] [17]

A probabilistic plan recognition algo- rithm based on plan tree grammars

[Geib and Goldman, 2009] Christopher W Geib and Robert P Goldman. A probabilistic plan recognition algo- rithm based on plan tree grammars. Artificial Intelligence, 173(11):1101–1132,

work page 2009